Abstract:
Riau University Library provides a variety of services for its visitors. One of the
services provided is book loan transaction service. Every day in the library there will be
dozens of book loan transactions and in general the transaction data is only used as an
archive without being used to generate information or patterns of book borrowing on a
data set. To find the pattern, it is necessary to analyze the data of book loan transactions.
One of the analyses that can be used is by data mining method of association rules using
algorithms (FP-Growth). In implementing FP-Growth method, we used data from book
loan transactions in Library of Riau University which consist of 18721 transactions and
the parameters used are minimum support 40% and minimum confidence 75%. FPGrowth
completion stages are to form FP-Tree and then fp-tree mining process by
forming conditional pattern base, forming FP-Tree and forming frequently pattern
growth. From these stages, the results obtained were 102 rules with several
combinations of items, namely a combination of 2 items, a combination of 3 items and a
combination of 4 items. In conclusion, the rules produced in this study are influenced by
the minimum value of support and minimum confidence, the greater the value, the more
accurate the rules will be.
